Buzzwords such as sprawl and smart growth have seeped into Washington politics, as Al Gore and other pols attempt to assemble a national consensus on land use policy. But anyone hoping to fashion a sensible program to deal with the byproducts of growth will have to get around a pretty big obstacle: After years of study, hardly anyone can agree just what the effects of sprawl, for good or for ill, really are.
For evidence, examine The Costs of Sprawl—Revisited, a new report sponsored by the Federal Transit Administration. In 268 pages, it reviews virtually all the serious research on sprawl done during the last several decades and attempts to distill all the areas of agreement it can find. In its words, it seeks "to break down the phenomenon of sprawl into its basic alleged impacts, both positive and negative, and to detail deliberately the strengths and weaknesses of each impact statement with specific citations from the literature." If you're looking for a clear picture, expect to be disappointed.
For example: While there is general agreement that the cost of infrastructure is linked to the density of new development, researchers are in far less accord as to how much this has to do with sprawl, and many make the very reasonable argument that subsequent waves of development will fill in enough areas to recoup the initial costs. How legitimate is this argument? It's hard to say, since hardly anyone's tried to tabulate the relevant data or, for that matter, to decide just how the data should be tabulated. That raises another point, which the study neglects: whether the values in question are too subjective to be quantified at all.
Usually, scholars can't arrive at even that much consensus. Consider the issue of public operating costs–that is, whether the splintered mini-governments that characterize suburban sprawl are "duplicative" and, thus, spend public money inefficiently. This is a popular contention within the smart growth crowd, but there turns out to be little real evidence for it. As the study notes, public services in large and dense areas "are more complex and individualized than those in smaller, more sparsely populated jurisdictions." Given that the smart-growthers haven't tried to measure the quality of the services in question, there isn't much one can say with certainty about the relation of public operating costs to sprawl.
Does the research concur on anything? Yes, but even here the interpretations often differ. Just about everyone agrees that the number of vehicle-miles traveled has been increasing. According to some scholars, however, commute times have nonetheless remained constant. Almost by definition, sprawl eliminates farmland and environmentally fragile territories. Just how much it eliminates, however, is again a point of dispute. And while sprawl is clearly linked to congestion, no one's sure whether the congestion's been getting better or worse. Or, rather, lots of people are sure; they're just sure of radically different things.
It gets worse. According to the report, few analyses of sprawl adequately define such basic terms as density, land consumption, and sprawl itself. Needless to say, this leads to some basic accounting problems. "Take, for instance, land consumption," the authors write. "Although it is a tautology that development consumes land, does a single-family home built on a 50-acre farm 'consume' all of those 50 acres? If it consumes only a fraction, on what basis is that fraction apportioned?"
The authors further comment that few scholars have undertaken any comprehensive empirical analyses of sprawl, that those who have done so have neglected important questions, and that most of the literature looks at only a limited time frame. They also note that "even the most current literature on sprawl tends to describe its attributes rather than quantify them." By the time they get to accusing scholars of using simplistic models, one gets the impression not only that we don't know much about suburban growth now but that we never will.
Chances are small, of course, that Al Gore's army of wonks will let that stand in their way.